TAPPI TIP-0402-29 Duplex Welding Qualification

TIP 0402-29
ISSUED - 2001
2001 TAPPI
The information and data contained in this document were
prepared by a technical committee of the Association. The
committee and the Association assume no liability or responsibility
in connection with the use of such information or data, including
but not limited to any liability under patent, copyright, or trade
secret laws. The user is responsible for determining that this
document is the most recent edition published.
Qualification of welding procedures for duplex

stainless steels
Scope
This TIP reviews the current state of knowledge regarding duplex stainless steels, particularly as it affects
their use in the pulp and paper industry. It addresses the definition of testing for qualification of weld procedures
that are both technically and economically effective. The discussion focuses on 2205 duplex stainless steel and on
digesters. However, the information will be generally applicable with appropriate modification for other duplex
stainless steels and to other vessels and processors built of duplex stainless steels within the pulp and paper industry.
Safety precautions
Any welding process requires appropriate safety precautions, techniques, and safety equipment (helmet,
shields, gloves, etc.). Normal manufacturing safety precautions apply. Some companies require welding permits.
Check with the appropriate department at your organization.
This TIP may require the use, disposal, or both of chemicals that may present serious hazards to humans.
Procedures for the handling of such substances are set forth on Material Safety Data Sheets that must be developed
by all manufacturers and importers of potentially hazardous chemicals and maintained by all distributors of
potentially hazardous chemicals. Prior to the use of this technical information paper, the user should determine
whether any of the chemicals to be used or disposed of are potentially hazardous and, if so, should follow strictly
the procedures specified by both the manufacturer, as well as local, state, and federal authorities for safe use and
disposal of these chemicals.
General guidelines for welding duplex stainless steels
It is assumed that the reader already has experience in welding of austenitic stainless steels such as Type
316L in pulp and paper mill applications. This section addresses some to commonly discussed welding
characteristics and procedures of the duplex stainless steels in terms of how they differ from austenitic stainless
steels. Addressing each of these features is essential for the design of technically and economically effective
welding procedures to be qualified.
The 2205 duplex stainless steel is far and away the most readily available grade with multiple producers of
virtually all product forms and good availability through common metal distribution channels. The issues regarding
welding of duplex stainless steels are similar for most grades so that a thorough discussion of 2205 is a sound basis
for minor modification to accommodate the other duplex grades, whether of greater or lesser alloy content.
TIP Category: Automatically Periodically Reviewed (Five-year review)

TAPPI
TIP 0402-29 Qualification of welding procedures for duplex stainless steels / 2
Differences between duplex and austenitic stainless steels
Duplex stainless steels are typically twice as strong as common austenitic stainless steels. The thermal
expansion of the duplex grades is between that of carbon steel and the austenitic stainless steels. The thermal
conductivity of the duplex stainless steels is also between that of carbon steels and the austenitic stainless steels.
When there are problems with welding of austenitic stainless steels, those problems are most frequently
associated with hot cracking of the weld metal itself. This hot cracking tendency is caused by the combination of
high thermal expansion and low thermal conductivity of austenite, but it is aggravated by fully or predominantly
austenitic solidification that occurs for the more highly alloyed austenitic grades. For the more common austenitic
stainless steels, hot cracking is minimized by adjusting the composition of the filler metal to provide a significant
ferrite content. For the more highly alloyed austenitic stainless steels where the use of a nickel-base filler metal is
necessary, austenitic solidification is unavoidable. In these cases these problems must be managed by minimizing
joint constraint and by low heat input, often requiring many passes to build up the weld.
Duplex stainless steels have good hot cracking resistance. Hot cracking of the duplex weld metal is seldom a
concern. The problems most typical of duplex stainless steels are associated with the HAZ, not with the weld metal.
The HAZ problems are not hot cracking but rather a loss of corrosion resistance and toughness, or of post-weld
cracking. To avoid these problems, the welding procedure should focus on minimizing total time at temperature
in the “red hot” range for the whole procedure rather than managing the heat input for any one pass. Experience has
shown that this approach can lead to procedures that are both technically and economically optimal.
The data in the appendix of ASTM A 923 suggest how rapidly intermetallic phases can precipitate to the
1
extent that corrosion resistance and toughness are significantly affected.
With this introduction in mind, it is possible to give some general guidelines for welding of duplex stainless
steels and then to apply this background and those guidelines to specific welding methods.
Selection of starting material
Duplex stainless steels are much more sensitive during welding to minor within-grade variations in chemistry
or processing than are austenitic stainless steels. For example, the importance of having sufficient nitrogen in the
duplex stainless steel base metal has been repeatedly emphasized. Air cooling of a plate, even when rapid, through
the 705 to 980° C (1300 to 1800° F) range will use up some of the “time on the clock” for the welder to complete
the weld before detrimental reactions occur. Similarly, if a plate is allowed to air cool into this range during transfer
to water quenching, that time is no longer available to the welder. The metallurgical condition of the material used
in actual fabrication should be the same quality with regard to composition and production practice, as the material
used to qualify the welding procedure. The specification of composition and testing for the starting material are
discussed in Section III and in Appendix B.
Cleaning before welding
The need to clean prior to welding applies to all stainless steels. The duplex stainless steels are sensitive to
contamination, and especially to moisture. The chemistries of the base metal and the filler metals have been
developed assuming no additional sources of contamination. Dirt, grease, oil, paint, and sources of moisture of any
sort will interfere with welding operations and adversely affect the corrosion resistance and mechanical properties
of the weldment. No amount of procedure qualification is effective if the material is not thoroughly clean before
welding.
Joint design
Duplex stainless steels require good joint preparation. For duplex stainless steels, a weld joint design must
facilitate full penetration and avoid autogenous (i.e., no filler metal) regions in the weld solidification. It is best to
machine rather than grind the weld edge preparation to provide uniformity of the land thickness or gap. When
grinding must be done, special attention should be given to uniformity of the weld preparation and the fit-up. Any
grinding burr should be removed to maintain complete fusion and penetration. For an austenitic stainless steel, a
skilled welder can overcome some deficiencies in joint preparation by manipulation of the torch. For a duplex
stainless steel, these techniques can cause a longer than expected exposure in the harmful temperature range, leading
to results outside of those of the qualified procedure.
3 /Qualification of welding procedures for duplex stainless steels TIP 0402-29
2
Examples of joint designs used with duplex stainless steels are shown in Figure 1. Other designs are
possible provided they assure full penetration welds and minimize the risk of burn-through.
Preheating
As a general rule, preheating of duplex stainless steel is not recommended because it slows the cooling of the
heat-affected zone. Preheating should not be a part of a procedure unless there is a specific justification.
Preheating may be beneficial when used to eliminate moisture from the steel as may occur in cold ambient
conditions or from overnight condensation. When preheating to remove moisture, the steel should be heated to
about 95° C (200° F) uniformly and only after the weld preparation has been cleaned.
Preheating may also be beneficial in those exceptional cases where there is a risk for forming a highly ferritic
HAZ because of very rapid quenching. Examples include welding a thin sheet to a plate, as with a liner to a vessel
or a tube to a tubesheet, or any very low heat input weld where there is exceedingly rapid cooling. (See FAQ 13.)
In such cases the heavier member should be heated to a temperature not to exceed the typical maximum interpass
temperature of 150° C (300° F).
Heat input and interpass temperature
Compared to austenitic stainless steels, duplex stainless steels can tolerate relatively high heat inputs. The
duplex solidification structure of the weld metal is resistant to hot cracking, much more so than that of highly
austenitic weld metals. Duplex stainless steels, with higher thermal conductivity and lower coefficient of thermal
expansion, do not create the same high intensity of local thermal stresses at the welds of austenitic stainless steels.
While it is prudent to avoid severe restraint, hot cracking is seldom a problem.
To avoid problems in the HAZ, the weld procedure should allow rapid (but not extreme) cooling of this
region. The temperature of the work piece is important because the plate itself provides the most effective cooling
of the HAZ. Typically, the maximum interpass temperature is limited to 150°C (300°F). That limitation should be
imposed when qualifying a weld procedure, and production welding should be monitored to assure that the interpass
temperature is no higher than that used in the qualification. Electronic temperature probes and thermocouples are
the preferred instruments for monitoring the interpass temperature. When a large amount of welding is to be
performed, planning the welding to provide enough time for cooling between passes is good, economical practice.
The size of the test piece used in qualifying a weld procedure may affect the cooling rate and the interpass
temperature. There is a risk that the test piece for qualification of a multipass weld procedure may come to a lower
interpass temperature than can be reasonably or economically achieved during actual fabrication. Therefore, the
qualification might not detect the loss of properties that can occur when a higher interpass temperature slows the
cooling and increases the time at temperature for the HAZ in actual practice. (See FAQ 5, 6.)
Postweld heat treatment
Postweld stress relief is not necessary or useful for duplex stainless steels. Unlike the L-grade austenitic
stainless steels, the duplex stainless steels are sensitive to even relatively short exposures to temperatures in the 300
to 1000° C (570 to 1800° F) range. Thermal stress relief in the 300 to 700° C (570-1300° F) range may cause
precipitation of alpha prime phase (“475° C (885° F) embrittlement”), causing a loss of toughness and corrosion
resistance. Stress relief in the range of 700 to 1000° C (1300 to 1800° F) leads to rapid precipitation of intermetallic
phases with moderate to severe loss of toughness and corrosion resistance. Any heat treatment of a duplex stainless
steel for whatever reason, should be a full solution anneal, meeting the minimum temperatures specified for the mill
product in the ASTM specifications, followed by water quenching. For 2205 that minimum temperature is 1040° C
(1900° F) in most cases. In areas where there has been little dilution of the higher nickel “matching” filler metal,
such as the 2209 that is used for the 2205 stainless steel, it may be necessary to increase the annealing temperature
to 1065° C (1950° F) to assure that the intermetallic phase is fully re-dissolved.
Some types of equipment manufactured from duplex stainless steel require a full anneal. For example, the
forming of large heads or the fabrication of some valve and pipe assemblies may require annealing. When there is a
full solution anneal and quench subsequent to welding, that heat treatment is a part of the welding procedure.
Annealing can restore the equilibrium phase balance and eliminate the problems associated with excessive ferrite
and intermetallic phases. If the common duplex filler metals are used, typically overalloyed with nickel, phase
balance in the fully annealed weld may shift toward austenite. Water quenching is essential after the final anneal,
but air cooling from intermediate thermal exposures, such as in hot forming, has been found to be practical and
economical. (See FAQ 12, 18.)
Phase balance in the weld
Modern duplex stainless steel mill products are balanced to have about 40-50% ferrite with the balance being
austenite. It is generally agreed that the characteristic benefits of duplex stainless steels (strength, toughness,
corrosion resistance, resistance to stress corrosion cracking) are achieved when there is at least 25% ferrite with the
balance austenite.
The ferrite in the weld metal is typically in the range of 25 to 60%. In some welding methods, particularly
those relying upon flux shielding, the phase balance of the filler has been adjusted toward more austenite to provide
improved toughness, offsetting the loss of toughness associated with oxygen pickup from the flux. There have been
no reports of problems associated with the ferrite contents at the lower end of this range, typically seen in SMAW
(shielded metal arc, or stick) or SAW (submerged arc) welds. (See FAQ 16, 17.)
Rapidly quenched autogenous welds, e.g., arc strikes, repair of arc strikes, small GTA repair welds, etc., tend
to have high ferrite, greater than 60%. Such welds can have low toughness and reduced corrosion resistance. (See
FAQ 4, 10, 13.)
Metallographic evaluation of the phase balance in the HAZ is an appropriate test for welding procedure
qualification. However, metallographic evaluation is not economically effective for evaluation of annealed mill
products or production welds. In the case of annealed mill products, the combination of narrow composition range
and tightly controlled heat treatment with rapid cooling assures a narrow range on the balance of austenite and
ferrite. Magnetic evaluation of the phase balance is, with a few limited exceptions, not effective. (See FAQ 17.)
Dissimilar metal welds
Duplex stainless steels can be welded to other duplex stainless steels, to austenitic stainless steels, and to
carbon and low alloy steels. Duplex stainless steel filler metals with increased nickel content relative to the base
metal are most frequently used to weld duplex stainless steels to other duplex grades. When welding duplex
stainless steels to austenitic grades, the austenitic filler metals with low carbon and a molybdenum content
intermediate between the two steels are typically used. AWS E309LMo/ER309LMo is frequently used for these
joints. The same filler metal or AWS E309L/ER309L is commonly used to join duplex stainless steels to carbon and
low alloy steels. Because most austenitic stainless steels have lower strength than the duplex grades, attention
should be given to whether or not strength is maintained when welded joints are made with austenitic filler metals.
When welding the highly alloyed austenitic stainless steels, nickel-base fillers are used. The nickel-base
filler metals are not normally used for duplex stainless steels, but if they are, they should be free of niobium
(columbium). Although not thoroughly documented, there have been reports that the ENiCrMo-3 filler (625) has
been less than fully satisfactory with duplex stainless steels, possibly because of interaction of the niobium from the
filler with the nitrogen from the duplex base metal.
Table I summarizes filler metals frequently used to weld duplex stainless steels to dissimilar metals. These
examples show the AWS bare wire designation (ER), but depending on the process, joint geometry and other
considerations, electrodes (AWS designation E), and flux-cored wire may be considered.
Table 1. Welding consumables used for dissimilar metal welding.

2304 2205 25 Cr Superduplex
2304 2304, ER2209 ER2209 ER2209 ER2209
2205 ER2209 ER2209 25Cr-10Ni-4Mo-N 25Cr-10Ni-4Mo-N
25 Cr ER2209 25Cr-10Ni-4Mo-N 25Cr-10Ni-4Mo-N 25Cr-10Ni-4Mo-N
Superduplex ER2209 25Cr-10Ni-4Mo-N 25Cr-10Ni-4Mo-N 25Cr-10Ni-4Mo-N
304 ER309LMo, ER2209 ER309LMo, ER2209 ER309LMo, ER2209 ER309LMo
316 ER309LMo, ER2209 ER309LMo, ER2209 ER309LMo, ER2209 ER309LMo, ER2209
Carbon steel ER309L ER309L ER309L ER309L
Low alloy steel
Welding procedure qualification
“Qualification” of welding procedures for duplex stainless steels must be considered in a broad sense, i.e.,
demonstration that the welding procedure that will be used in fabrication will not lead to unacceptable loss of
engineering properties, especially toughness and corrosion resistance. For other types of stainless steels,
qualification testing for weld procedures is fairly simple, with only a limited amount of testing to qualify a material,
filler metal, and weld method. With hardness tests and bend tests, looking for martensite and hot cracking,
respectively, these qualification tests reflect long experience for what can go wrong when welding ferritic,
martensitic or austenitic steels. Duplex stainless steels are unlikely to have difficulty meeting these requirements,
because these standard tests are unlikely to find intermetallic phases or excessive ferrite, the most likely problems
for duplex stainless steels. A bend test may still be useful and economical, but it is not conservative in the sense of
always detecting problems if present. Because of the limitation on total time at temperature for the HAZ, the
properties of duplex grades will be sensitive to section thickness and details of actual welding practice. So the
qualification of procedures for duplex stainless steels is specific to particular geometries of welding, much more so
than for austenitic stainless steels.
It would be conservative to qualify a weld procedure for every thickness, geometry, and method of welding
because minor differences in setup may be significant in the results achieved in production. However, the complex
nature of actual constructions makes such testing costly. Savings are achieved by qualifying the procedures
(defined by section, filler, and method) determined to be the most demanding on the duplex stainless steel. It is
prudent to test welds in the most critical joints in a construction, even when those joints might be exempt from
testing of production welds under a strict reading of the ASME UHA 51. The temperature for toughness tests will
depend on whether the test is to check the metallurgical condition of a mill product (as in the use of ASTM A 923 to
detect intermetallic phases) or to demonstrate the suitability for use of a construction (as in the use of ASME UHA
51). (See FAQ 15, 16.)
Applicable welding methods
Second-generation (nitrogen-alloyed) duplex stainless steels saw rapid development in the early 1980s. With
only limited understanding of the formation of intermetallic phases, early views of welding duplex grades focused
on limiting heat input, possibly because this approach is what is typically applied to special austenitic grades. With
such severe limitations on heat input, many of the more economical welding methods with high deposition rates,
such as submerged arc welding, were thought to be inappropriate for the duplex stainless steels. However, the final
properties of the duplex stainless steels are of such interest that much effort was directed to learning how to use the
more economical processes. Now virtually all welding processes, except for oxyacetylene with its associated
carbon contamination of the weld, are applied to duplex stainless steels. Gas Tungsten Arc Welding (GTAW), Gas
Metal Arc Welding (GMAW), Shielded Metal Arc Welding (SMAW), Flux Core Arc Welding (FCAW),
Submerged Arc Welding (SAW), and Plasma Arc Welding (PAW) have all seen practical application. Electric
Resistance Welding and Electron Beam Welding, although much less common, have also been qualified and used in
particular fabrications.
There are important differences among the welding procedures. For example, the decision to use a flux-
shielded weld and selection of flux for that weld affect toughness. (See FAQ 15, 16.)
Issues to be addressed
Starting material. It is essential that the welding procedure qualification be performed on material that is
essentially identical in specification and actual production to the material that will be used in fabrication. For
example, if material used in the qualification is of significantly higher nitrogen or has been quenched much more
rapidly than the material that is used in fabrication, then one cannot rely upon the use of the qualified procedure to
produce a weld of acceptable properties.
For the starting material, the ASTM specifications that cover duplex stainless steels require testing of each
heat lot of material. The ASTM definition of heat lot for a continuous heat treatment system permits all material of
a given heat and size, processed within the same operating period and under the same time and temperature, to be
represented by a single sampling. However, ASTM specifications do not address the possibility for an individual
plate of duplex stainless steel to dwell in the temperature range below the specified minimum annealing
temperature, as may occur during “rapid air cooling” or in the transfer of product from the furnace to the quenching
operation. The time for the quenching operation will vary from one producer to another and will vary by size range
for an individual producer. Therefore, for the most critical applications, the purchaser may choose to require testing
of each piece even within a heat lot.
In the ASTM specifications, the S32205 version of 2205 has narrower (the upper half) ranges on chromium,
molybdenum, and nitrogen than does S31803. These ranges reflect the experience accumulated over almost twenty
years that consistent, favorable results are obtained by tight control of these elements. Furthermore, the heat
treatment requirements for S32205 include water quenching except for coiled product heat treated in a continuous
annealing line. This exception to water quenching is applicable to the thin product, sheet, where consistent rapid air
cooling is effective and essential to operation. Imposing a similar restriction on the quenching required for other
duplex stainless steels would be an appropriate and conservative requirement.
All duplex stainless steels should be ordered as complying with ASTM A 923. Not all duplex grades are
presently listed in A 923, but at least Method A can be applied as an acceptance test. The acceptance criteria for
Methods B and C for grades not specifically covered are subject to agreement between producer and user. For
S32205 (and S31803), specification of both Methods B and C may not be necessary, incurring extra cost, but is
conservative. A 923 permits the use of Method A as a screening test, where acceptance replaces the need to
complete Methods B and C. The user may require reporting of actual test results for Methods B and C as a way of
removing the permission to use Method A for screening. As noted below, the impact tests for the ASME
requirements may be done at the more stringent condition of A 923 in order to save on testing costs.
The ASME toughness testing addresses the suitability for use of a particular construction, and the
3
requirements depend on section thickness and minimum design metal temperature. The ASME specification for
duplex stainless steels, when applicable to a particular construction, require the more common set of three
specimens for the Charpy test, and use lateral expansion rather than energy absorption as the acceptance criterion.
The impact test requirements of the ASME specification are less stringent than those of A 923, unless the minimum
design metal temperature is below –40° C (-40° F). It is possible that the ASME required testing would not detect
the early stages of formation of intermetallic phases. But as noted in ASTM A 240, performing the ASME tests at
the more stringent test temperature of A 923, with measurements of both impact energy and lateral expansion, is
permitted and allows for cost saving by avoiding redundant testing. (See FAQ 15, 16).
Hardness tests of the starting material are of little value with respect to the determination that heat treatment
has retained toughness and corrosion resistance through the elimination of intermetallic phases, but fortunately
hardness tests are also of little cost. If hardness testing is required, as is the case for most duplex grades in the
ASTM specifications, the maximum should be that listed in the specifications.
Sampling. It is important to qualify both the heat-affected zone and the weld metal itself for a duplex
stainless steel. However, the testing procedure and acceptance criteria are likely to differ for the HAZ and the weld.
In the case of austenitic stainless steels, there are seldom problems associated with the heat-affected zone, other that
the possibility of sensitization, relatively unusual for the low carbon versions of these grades.
There are several situations in which it is necessary to qualify the welding procedure for a duplex stainless
steel at different positions in the weld with regard to the thickness. For example, for a welding procedure calling for
a GTA root pass followed by filling with another weld method, it is appropriate to test the weld both in the area of
the root and at one or more levels within the filler passes. In multi-pass welds, each subsequent pass can have
significant effects on the condition of the prior passes, such as altering phase balance, causing formation of
secondary austenite, or causing precipitation of non-metallic or intermetallic phases.
Acceptance criteria. The appropriate acceptance criteria for welds, applicable to as-welded fabrications,
vary widely from those applicable to annealed mill products, especially with regard to toughness measurements for
flux-shielded welds. (See FAQ 9, 15, 16.)
Guidelines for testing for procedure qualification
For the qualification of the welding procedure, it is recommended that the following tests be performed for
each significant geometry and thickness. Some judgment must be applied with regard to the differences in the
proposed procedures with respect to total time at temperature and the rate of quenching that will be seen for the last
pass of welding. For example, it is convenient but possibly unrealistic to weld plates in the flat, downhand position
to qualify a procedure if it is known that the actual welds will be made with less than perfect fit-up or with
substantially out-of-position welding.
It is common for a skilled welder to overcome minor deviations in fit-up with welding technique and local
rework, and these may lead to extended time at temperature for the HAZ. Accordingly, it is prudent to qualify
procedures for reasonably anticipated repairs or rework of welds. For example, when welds are to be radiographed,
it is obviously the intention that any defects found will be repaired. Therefore, it is appropriate and economical to
qualify a repair procedure for one or two successive repairs to be performed on piece that has been welded by the
qualified fabrication welding procedure. Qualification of the major repairs should use the same tests as the
qualification of the procedure for the fabrication weld. For minor repairs, it may be sufficient to perform only
limited confirmation tests, such as the tests of ASME UHA-51 or ASTM A 923 Method C.
For the welding procedure qualification, the following materials and tests are typically specified:
Starting material (mill products)
1. Material should meet the appropriate ASTM product form specification (or ASME specification, as noted
below, for Code-qualified construction). For 2205, the S32205 should be specified. For other duplex stainless
steels it is recommended that the listed composition be accepted, but with consideration for further limits on
composition. For those grades with nitrogen content range (max-min) in excess of 0.10%, chromium content
range in excess of 1.0%, and molybdenum content range in excess of 0.50%, it is appropriate to further restrict
the nitrogen, chromium, and molybdenum to ensure a uniformity of response.
2. The requirements for quenching after heat treatment applicable to S32205 should be specified as applicable to
the other grades of duplex stainless steels.
3. The user may wish to specify no weld repair of mill products without specific permission.
4. The material should be tested in accordance with ASTM A 923, Methods A, B, and C, with requirement to
report the test values obtained in Methods B and C. Methods A and C, if specified, shall apply at all
thicknesses. Method B, if specified, shall apply at all thicknesses, 2.5 mm (0.1 inch) or greater. (Note possible
cost savings by using same test to meet ASME requirements when applicable.)
5. The phase balance of the mill product shall be in the range of 35 to 60% ferrite, balance austenite, with testing
by automated image analysis being permitted.
6. For ASME Code construction, material shall be specified as meeting the appropriate ASME product form
specification. For 2205, S31803 should be specified with all specific further restrictions as applicable to
S32205 (often achieved by ordering dual certified S31803/S32205).
7. For ASME Code construction, Charpy impact tests (three specimens) shall be performed at the minimum
material design temperature for the required section thicknesses, meeting the acceptance criteria for lateral
expansion, as defined in ASME UHA 51. These tests are permitted to be performed at a lower temperature,
such as that in A 923, to reduce cost of redundant testing.
Welding procedure qualification testing
1. Bare wire filler metal shall be selected in accordance with AWS specifications. Flux-shielded filler shall be
selected in accordance with AWS specifications and, because the proprietary formulations of these fluxes may
significantly affect the properties of the resulting weld, with the filler manufacturer.
2. A tensile test transverse to the weld shall meet the requirements of the base material. Face and root bends of
the weld shall be performed in accordance with the requirements for the base material.
3. Charpy impacts tests shall be performed at –40° C (-40° F) on the weld and on the HAZ for material having
thickness of > 3/16 inch, and at lesser thicknesses at the minimum metal design temperature when required by
ASME UHA 51. Care should be taken in locating the notch for the HAZ Charpy specimens to encompass the
HAZ and the fusion line of the weld.
For Code applications, the Charpy test shall be performed in accordance with the requirements of UHA 51. It
is permitted to perform these tests at a temperature lower than the minimum design metal temperature.
For non-Code applications, a Charpy test in accordance with ASTM A 923 Method B is permitted. However,
A 923 was developed for annealed mill products. The acceptance criterion of 40 ft-lb at –40° C (-40° F) is not
applicable to the weld metal, in the sense that a lower value of the impact energy may not be attributable to
intermetallic phases. The acceptance criterion for annealed mill products may not be applicable to the HAZ in all
cases. Therefore, if an impact energy is to be used for acceptance of the weld and HAZ, the acceptance criterion
must be reduced appropriately from the level required of the annealed mill product. The impact energy for flux-
shielded welds may have a shelf energy of as little as 20 ft-lb. It is unreasonable to require that the weld must have
the same high impact strength as the annealed base metal, and it is uneconomical to not accept the welding methods
that regularly produce a lesser but still acceptable impact energy. (See FAQ 15, 16.)
4. The user may elect to perform a corrosion test to confirm the absence of detrimental intermetallic phases or
non-metallic phases. The test specimen should encompass weld metal, HAZ, and base metal, but it is not
necessary to encompass the entirety of the weldment for larger samples. The acceptance criteria for A 923
were selected as applicable to annealed mill products. The edges of the test specimen shall be ground smooth,
but not chemically pickled. It is appropriate to slightly reduce the required temperature of testing when an
unannealed weld is present, e.g., to 22° C for S32205. Corrosion on the sample edges shall not be ignored as
permitted in ASTM G 48. (See FAQ 8, 9.)
5. Penetrant tests shall be performed on the qualification coupons when required by ASME procedures.
6. The phase balance shall be determined metallographically for the weld and heat affected zone. For welds, the
ferrite shall be at least 25% for the flux-shielded welds, and at least 30% for gas-shielded welds. (See FAQ
17.) For the HAZ, it is desired that the ferrite shall be within the range specified for the starting material.
However, it is often difficult to measure the phase balance precisely because of the typically small and variable
size of HAZ and the variation of ferrite content within the HAZ. In no case shall there be a continuous layer
of fully ferritic material surrounding the weld.
Testing for production weld test plates
1. Material used in the fabrication shall conform to the same specification as used for the material used in the weld
qualification.
2. Where the welding geometry permits, and with a reasonable frequency for such opportunities, runout tabs of
the same heat lot of plate shall be attached to the fabrication plate and the weld continued into the tab without
change of procedure. The runout tab shall be evaluated in accordance with the requirements of the ASME, if
applicable. In the most conservative approach, a sampling of the runout tabs shall be evaluated using the same
tests and acceptance criteria as used in the applicable weld procedure qualification.
3. The runout tab shall be of sufficient size to allow the conditions of welding to be maintained. Sizes of 150 x
150 x t mm (6 x 6 x t in.) to 300 x 300 x t mm (12 x 12 x t in.) are typically used.
4. The heat input shall be measured and reported as a monitoring of the uniformity of procedure. The interpass
temperature shall not exceed 150° C (300° F) as measured by a contact thermocouple.
5. Welding that deviates from a qualified procedure shall be reported and documented. The user may require
qualification of the actual procedure before accepting the weld.
Training of welders and inspectors
Because the thermal expansion and thermal conductivity of duplex stainless steel differs markedly for those
of austenitic stainless steel, it is useful for welders to have the opportunity to make a number of trial welds for
different geometries and qualified procedures.
It is essential that individual welders and inspectors of the fabricator appreciate the risk to toughness and
corrosion resistance of any deviation from the qualified procedures. Simply stated, any deviation from the qualified
procedure is more likely to be harmful and is likely to be more harmful that a deviation from procedure for an
austenitic stainless steel. While there are some deviations that can be tolerated, it must also be appreciated that
seemingly minor deviations can have very large effects. One example is that of a conscientious welder who sought
to apply a very small GTAW smoothing pass, not included in the qualified procedure, after the weld had thoroughly
cooled. The result was that the solidification of that pass produced nearly 100% ferrite in the weld zone, with
significant loss of corrosion resistance.
Other examples of actions that can be more harmful to duplex stainless steels than to austenitic stainless
steels include spatter, arc strikes, and small surface repairs. For duplex stainless steels they are likely to produce
small regions of highly ferritic structure, with the possibly associated loss of corrosion resistance.
Maintaining dry welding filler materials, and particularly those with flux shielding, is especially important for
duplex stainless steels. Practice should begin with oven storage and “hot-box” holding areas. It is also observed
that there is greater sensitivity to flux formulation, possibly leading to brand-specific response within a qualified
procedure.
It is essential that the welders and inspectors of any subcontractor be adequately trained in the issues
discussed above. Similarly, the welders and maintenance engineers of the user/operator of the equipment should be
4
trained in the technology of welding duplex stainless steels. (See TAPPI TIP 0402-23) (See FAQ 3, 4, 10, 13,
14.)
Summary
1. The duplex stainless steels are seeing substantial and diverse applications within the pulp and paper industry, as
appropriate to their excellent mechanical properties and corrosion resistance, and their economy.
2. A full range of welding practices has been successfully applied to duplex stainless steels.
3. The qualification of welding procedures requires testing of toughness and corrosion resistance to demonstrate
the absence of excessive ferrite or the absence of intermetallic phases.
4. The quality of starting material is an important issue in producing welds with consistent and acceptable
properties.
5. Total time at temperature for the HAZ is one of the most important control variables for the welding of duplex
stainless steel.
6. Repairs during fabrication or future maintenance must be considered and qualified as a complete procedure,
extending the originally qualified procedure.
Keywords
Digesters, Stainless Steel, Welding
Additional information
Effective date of issue: June 22. 2001.

Working Group Members:
Ralph Davison, Chairman, Technical Marketing Resources

J-P Audouard, Creusot-Loire Industrie
Don Bardsley, Beloit Corp.
Alan Castillo, International Paper
Bob Charlton, Levelton Engineering Ltd.
Gary Coates, Nickel Development Institute
David Crowe, Union Camp
Harry Dykstra, Bacon Donaldson Consulting Eng
Andy Garner, PAPRICAN
Richard C. Garvin, Creusot-Marrel Inc.
Margaret Gorog, Weyerhaeuser
John Grocki, Creusot-Loire Industries
Barry Heuer, Nooter Fabricators Inc.
Wendi L. Kaiser, Consulting Engineering Services
Steve Lukezich, Mead Central Research
Greg Makar, Westvaco
Steve Meiley, Champion International
Geir Moe, Rasco Specialty Metals
Max Moskal, Mechanical & Materials Engg
Alex Nadezhdin, Noranda Technical Centre
Subhash Pati, International Paper
Craig Reid, Bacon Donaldson Consulting Eng
Sandy Sharp, Westvaco
Doug Sherman, Mechanical & Materials Engineering
Tom Sprader, Thos P. Sprader & Associates
Arthur Tuthill, NiDI
Angela Wensley, Bacon Donaldson Consulting Eng
Paul Whitcraft, Rolled Alloys
Jim Willis, Mead Central Research
Barry Woit, Canspec
References
1. 1999 Annual Book of Standards, vol 01.03, ASTM, West Conshohocken, PA, 1999
2. TIP 0402-23, TAPPI, Atlanta, GA, 1998
3. ASME Boiler and Pressure Vessel Code, ASME, New York, NY, 1998
Appendix A. Testing requirements applicable to duplex stainless steels

Rationale for selecting test requirements
Because duplex stainless steels are still “new” to many users, it is common for specifications of duplex
stainless steels to call for testing requirements beyond those of the standard specifications. Examples of these
special requirements include: modified chemistries; tensile tests; bend tests; hardness tests; impact tests; phase
balance; metallographic evaluation; magnetic testing; and corrosion testing. Extra testing is not usually harmful, but
it may be expensive and it may provide a false sense of security if the testing does not meaningfully address the
perceived problem. However, some of the tests that have been considered and even required in specifications of
duplex stainless steels, such as a maximum hardness, have actually been harmful because the test requirements
could be met only by deviating from optimal chemical composition or processing.
It is important to differentiate tests applicable to mill products from those applicable to qualification of
fabrication procedures and from those to determine fitness-for-service in specific applications.
Testing requirements should be evaluated according to the following criteria:
• Conservative capability to address the property of interest;

• Definition of test sample, location and testing frequency;
• Applicability to all product forms and constructions of interest; and
• Cost.
The meaning of “conservative capability” in this case is that the test will always identify or disqualify
unacceptable material but will not too frequently disqualify acceptable material.
With increasing experience, some of these special requirements that have been applied to duplex stainless
steels are being set aside while others are finding their way into standard specifications. For example, ASTM A 923
was developed for the testing of duplex stainless steel mill products for the absence of detrimental intermetallic
phases. The following discussion reviews candidate tests for technically conservative results and for cost
effectiveness in establishing the properties of duplex stainless steels.
Chemical composition requirements
The properties of duplex grades are highly sensitive to variations in their chemical composition, particularly
the chromium, molybdenum, nickel, and nitrogen, among the specified elements. (A specified element is one
controlled to a specified range, maximum, or minimum, in accordance with the requirements of the ordering
specification.) The interactions of the alloy element additions are quite complex with respect to phase balance,
formation of non-metallics and intermetallics, and response to heat treatment and welding. After a decade, some
clear guidelines have evolved, particularly for 2205, as embodied in S32205. As shown in Table 3, the chemical
limits for 2205 originally were broad, as shown for S31803. With experience, a more appropriate balance of
elements has emerged. Initially 2205 was produced with chromium and molybdenum at their nominal levels, 22%
and 3%, respectively, but these began to slip toward their minimums. The new S32205 enforced the nominal limits
for chromium and molybdenum. The importance of nitrogen is well known, and it is readily controlled within the
upper half of the original range. Nickel is the least critical of these elements because it does not have an effect on
chloride pitting resistance; however, nickel does influence phase balance.
A little knowledge about chemical requirements can be a dangerous thing. Because end users are aware of
the pitting resistance equivalent (PRE), in some cases, they have set minimum values of PRE that are equal to or
higher than that typically produced for a particular grade. Specifying a non-standard chemical composition may
result in loss of properties other than corrosion resistance, loss of access to service center inventory not meeting the
special chemistry, and possibly increased costs of production. The PRE equation contains chromium, molybdenum,
and nitrogen, but these elements are not metallurgical substitutes for one another. The PRE is a relationship derived
from a statistical regression analysis. To properly use the resulting relationships, it is important to recognize the
assumptions that went into the analysis. In the original data, chromium, molybdenum, and nitrogen were not varied
independently. Any interpretation that the relationship asserts an independent, linear effect of each of these
elements on pitting resistance is not correct. The PRE equation for duplex stainless steels is valid only for balanced
alloys. “Balanced” means that chromium, molybdenum, and nitrogen are jointly coordinated to provide a duplex
steel with only two phases having optimal mechanical properties, phase stability, ease of processing, response to
heat treatment, weldability, and corrosion resistance. Therefore, the PRE may be considered a useful relationship
for ranking the chloride pitting resistance of duplex grades that have been optimized in composition, but not for
adjusting the composition within a particular grade without regard to phase balance. Unbalancing a composition
within a grade to meet a PRE goal may even diminish corrosion resistance if the effects on phase balance and
stability are not considered. It is recommended that the specification of duplex stainless steel chemistries be based
only on the standard chemical compositions, such as S32205 for 2205.
Tensile testing requirements
The formation of intermetallic phases, detrimental to corrosion resistance and toughness, may increase the
strength of duplex stainless steels. However, imposing a maximum on the strength of the duplex stainless steels is
not an effective means of controlling the presence of intermetallics. The duplex stainless steels are strong in the
annealed condition and they strengthen rapidly in response to light deformation as encountered in processing such
as straightening or forming. Higher nitrogen, which retards the formation of intermetallics, also increases strength.
High strength from work hardening or from higher nitrogen generally represents no engineering disadvantage.
Originally ASTM A 815 imposed a tensile strength maximum on 2205, 2507, and S32760 fittings. The most
probable reason for these limits was that the specification was originally written for martensitic stainless steels
where high strength is an indicator of untempered martensite. Faced with such limits for a duplex stainless steel, the
only option to reduce strength is to limit the nitrogen content, an action with undesirable side effects. The upper
limit on tensile strength has been eliminated for 2205 but no action has been taken yet on the other two duplex
grades. Specification of maximum strength for a duplex stainless steel is not an appropriate limitation for best
quality.
Elongation is a different matter. Austenitic stainless steel flat-rolled products are usually specified with a
minimum elongation of 40%, and it is not unusual for them to exhibit elongations of 50% or higher. Duplex
stainless steels have a specified minimum elongation of 15 to 25% depending on the grade. Actual elongations are
typically 30% but seldom exceed 35%. Leveling and straightening operations in duplex stainless steels exhaust a
larger portion of fully annealed elongation than for austenitic grades. For pipe and fittings, the flattening operation
to produce a flat tensile specimen may use a greater portion of the potential elongation than for austenitics. It is
particularly important to take into account the reduced elongation in a formed, as-welded part such as welded
duplex pipe manufactured without subsequent annealing, as may be specified with ASTM A 928. It is unlikely that
standard, room temperature tensile tests are sufficiently sensitive to detect intermetallics at the small volume
fraction (on the order of 1 vol. pct.) known to decrease corrosion resistance and impact toughness. Therefore, while
achieving the specified minimum elongation in a mill product is important for enabling further fabrication,
elongation itself does not effectively measure or control potential deleterious effects of intermetallic phase
formation.
Bend test requirements
Bend tests are not typically required for duplex stainless steel mill products. However, if they are, the
minimum radius for a “standard” bend test must be adjusted to reflect the lower total elongation characteristic of
duplex stainless steels even in the fully annealed condition. Because duplex stainless steels are so strong, bend tests
may not be convenient to perform, especially for heavier sections. A bend test is highly effective in detecting
cracking that may have occurred during rolling, e.g., hot tearing that may have resulted from rolling at too high a
temperature. A bend test may also be effective at detecting pervasive precipitation of intermetallic phase, e.g., one
that is associated with heat treatment rather than welding. Because the positioning of the bend may miss a critical
region or the remainder of the unaffected material may be capable of sustaining the deformation, a bend test may
not be sufficiently sensitive to detect intermetallics formed internally such as in a weld heat-affected zone.
Therefore, bend tests are a relatively low cost test for qualification of a mill product, and may be applied to
qualification of a fabrication procedure but may not be sufficient when applied to localized problems relating to
welding.
Hardness requirements
Because many of the early applications of duplex stainless steel were in the oil and gas industry and because
this industry is accustomed to dealing with steels and stainless steels that are capable of martensitic transformation,
many of the earlier specifications for duplex stainless steels required material not to exceed a hardness maximum.
This hardness maximum is correct for controlling martensite, but not for controlling duplex stainless steel. The user
would like to encourage the presence of nitrogen, but nitrogen increases both the yield stress and the rate of work
hardening. Consequently, any maximum hardness requirement works against obtaining the optimal duplex stainless
steel chemistry.
When hardness is specified, it is also important to specify where hardness shall be measured. Because the
end user requires flat or straight products, the steel producer applies mechanical straightening, for example, by
rolling, flexing, pressing, or stretching. The surface of the steel is hardened. It is most economical to measure the
hardness at the surface. If the hardness exceeds a specified maximum, one should have the option of measuring
below the surface, e.g., at quarter-thickness in flat-rolled products or half-radius in round bar or forgings, with the
option to meet that same hardness limit on the surface. This is a conservative approach because one should expect
that the surface will not be softer than the middle of the material.
Mircohardness measurements have been shown to be capable of detecting 475°C (885°F) embrittlement in
the ferrite phase of duplex stainless steels. However, a microhardness measurement samples only a very small
volume of material and can easily miss a detrimental condition such as intermetallic phase precipitation. Even
hardness measurements which sample a larger volume of material, such as Rockwell B and C, are not an appropriate
test to unambiguously detect the formation of intermetallic phases because high hardness values may also be a result
of thermomechanical processing. A metallographic examination of the cross section of a piece of steel can reveal
far more about its condition than any hardness measurement.
Impact toughness requirements
Duplex stainless steels are exceedingly tough materials. Mill plates typically have impact toughness levels
that approach or exceed 225 J (165 ft.-lb.) for longitudinal samples (long axis parallel to the rolling direction, notch
axis perpendicular to the rolling face) tested at -40°C (-40°F). The ductile-to-brittle transition is not sharp as it is in
a ferritic stainless steel, but rather exhibits a moderate slope from the upper shelf energy to about 50 to 75 J (35 to
55 ft.-lb.) at about -100°C (-150°F). The presence of the austenite in the duplex structure, when that phase is
unaffected by intermetallics, assures some toughness even down to low temperatures. The toughness of wrought
duplex stainless steels is strongly anisotropic. Because the product is rolled in the duplex temperature range, the
duplex structure is elongated. For plate, depending on the extent of cross rolling, the transverse Charpy sample
(long axis transverse to the rolling direction, notch axis perpendicular to the rolling face) exhibits one half to two
thirds of the shelf energy of the longitudinal sample. Although not widely documented, it is estimated that the
through-thickness Charpy specimen (long axis perpendicular to the rolling face) has a shelf energy that is about half
of that of the transverse specimens. But even in this least tough direction, the impact energy is still higher that the
toughness that has been considered acceptable for large pressure-containing fabrications in carbon steel.
Toughness tests for intermetallic phases (ASTM A 923, Method B). ASTM A 923 Method B uses the
Charpy test as an acceptance procedure for duplex stainless steel mill products to demonstrate the absence of
detrimental intermetallic phases. It uses a single specimen rather than the set of three for the typical Charpy test
because the A 923 test uses the absence of embrittlement as an indicator of the absence of the embrittling
intermetallic phases. At present, only S31803 and S32205 have an acceptance criterion. The acceptance criterion is
54 J (40 ft.-lb.) for a full-size Charpy, longitudinal impact specimen, tested at -40°C (-40°F) with a linear,
proportional reduction of the required minimum impact energy for subsize specimens. A 923 permits any specimen
orientation provided that the energy achieved is equal to that required for the longitudinal specimen, known to be
the toughest orientation. Testing transverse specimens is often economical for the cutting of specimens with
minimal loss of useful material, and it is conservative if the same acceptance criterion is met.
The acceptance criterion of 54 J (40 ft.-lb.) at -40°C (-40°F) is high relative to toughness requirements in
common fabrications, but it is appropriately conservative for detection of intermetallics in an annealed mill product.
This test was not intended to qualify toughness for welded fabrications. As stated in the specification, A 923 was
written for the qualification of mill products and may not be directly applicable to weld qualifications or to weld
fabrications. ASTM A 923 addresses, with a few exceptions, the occurrence of intermetallic phases in annealed mill
products.
In mill products, intermetallic phases may occur in the “centerline” of the product as a remnant of
solidification segregation. Method B is not especially effective in detecting centerline intermetallic phases.
Alternatively, intermetallic phase may occur pervasively as a result of excessive exposure in the 700-980°C (1300-
1800°F) range, resulting either from improper annealing temperature (failure to dissolve existing intermetallics or
creating more intermetallics in the annealing process itself) or from an inadequate quench rate allowing too much
time in the intermetallic formation temperature range during cooling. Method B is particularly effective at
detecting pervasive intermetallic phases in these cases.
Toughness tests for ASME requirements for service qualification. The ASME Boiler and Pressure Vessel
Code has added new requirements regarding the toughness of duplex stainless steels in Section VIII, Division 1,
Paragraph UHA-51. These requirements address the toughness of the mill product, the qualification of welding
procedures, and the qualification of actual fabrications in some cases, using Charpy tests at the minimum design
metal temperature (MDMT). In 1997 ASME requested ASTM to provide for supplemental requirements that could
be specified when ordering SA-240 duplex stainless steel plate. Action by ASTM is likely to require about one
year. The ASME requirements are for sets of three specimens, evaluated by lateral expansion rather than impact
energy. In the interest of minimizing costs, it may be possible to use -40°C (-40°F) as a conservative approach to
MDMT’s of -40°C or higher, and use the same mill tests to satisfy both the ASME Code requirement and the A 923
Method B requirement when Code qualified material is specified. Once the supplemental requirements have been
adopted in ASTM and ASME specifications, it is likely that most duplex stainless steel mill products will be
produced to these requirements to facilitate the use of service center inventories for ASME Code fabrications.
Toughness tests for weld qualification. For the qualification of welding, whether qualification of welding
procedures or actual fabrication welds, e.g., testing run-out tabs, the Charpy test must be carefully specified. The
results may be sensitive to the orientation of the mill product being welded, to the weld procedure, and to the precise
location of the notch. If the notch is not placed precisely in the region of the heat-affected zone (HAZ) with the
longest total time exposure in the 700-980°C (1300-1800°F) range, this test may not effectively detect the
occurrence of intermetallic phases resulting from welding. Because even a slight misalignment of the notch may
miss a problem, the Charpy test is not necessarily conservative in its evaluation. The acceptance criterion for a
Charpy test of the HAZ of a weld or any other steel heated but not remelted, e.g., a region warmed to facilitate
forming, is typically the same as for the mill product.
In using the Charpy test for qualifying a welding procedure, it is important to recognize that the results may
depend on the effectiveness of the quenching of the plate at the mill. The formation of intermetallics is controlled
by cumulative time in the 700-980°C (1300-1800°F) range, including both the time of cooling from solution
annealing at the mill and the welding time. So a plate that has been less effectively quenched from above 1050°C
(1925°F), e.g., it remained in the 925°-980°C (1700-1800°F) range for one or two minutes while being transferred
from the furnace to the quench as compared to a plate that was above 1050°C (1925°F) until quenched, may
produce a less favorable result after the time-temperature exposure in welding. It is essential that the plate used for
the actual fabrication be quenched as effectively as the plate that was used for the weld procedure qualification.
This issue can be addressed to some degree for 2205 by specifying S32205 rather that S31803 because the ASTM
specification requires water quenching for S32205 not processed in a continuous anneal and pickle line, rather that
the more common “water quenched or rapidly cooled by other means.” For other grades of duplex stainless steel, it
may be prudent to impose the specific requirement for water quenching from the annealing temperature as part of
the purchase order.
There is an important difference between the use of the Charpy test in A 923 and the more common use of
the Charpy test as test for fitness for service. The A 923 Practice B is an acceptance test for mill products. It is
based on a loss of toughness that is correlated with the appearance of the intermetallic phases metallographically
and with the associated measurable loss in corrosion resistance. The absolute value of the toughness, 54 J (40 ft.-lb.)
at -40°C (-40°F) for the mill product, is not as critical as the fact that the loss of toughness demonstrates that the
precipitation of intermetallic phases has occurred.
For weld metal, the use of the Charpy test is different because the shelf energy of the weld filler material is
expected to be significantly lower than the toughness of the base material. That is hardly an unusual observation for
a weld, but the differences between the filler metal and the base metal seem rather large for duplex stainless steels in
comparison with the differences typical for austenitic materials. Welding processes involving flux shielding, such
as shielded metal arc welding (SMAW) or sub-arc welding (SAW) produce significantly lower weld metal impact
energy than in the base metal, probably because of oxygen pick-up in the weld metal. If there are toughness
problems with the weld metal, they are more likely to be related to the method of welding or to excessively ferritic
structures than to the formation of intermetallic phases. The excessively ferritic structure may be at least partially
remedied by modification of the welding procedure. For the qualification of welding procedures, the Charpy test
result is important in an absolute sense, indicating a performance characteristic of the construction and not just the
presence or absence of a particular feature of the structure. Therefore, the appropriate acceptance criterion for
toughness of a weld, whether energy absorption or lateral expansion, is that set by fabrication codes such as the new
ASME requirements.
Phase balance requirements
When the second generation duplex stainless steels were being introduced, the determination of the phase
balance was given a lot of attention. Possibly it was a reflection of the experience with first generation duplex
stainless steels where, with the absence of nitrogen additions, there was a tendency to form excessive ferrite in the
HAZ after welding. This reduced both toughness and corrosion resistance. Possibly, it was just the comfort of
being able to measure something that would yield an understandable number to give a feeling of confidence
regarding these new, very strong but at that time unfamiliar grades.
It is generally accepted that the definition of a duplex stainless steel is a microstructure of austenite and
ferrite with each phase accounting for about 30 to 70 percent volume fraction. The precise defining volume
fractions may be debated, but the phase balance is typically spoken about as being “about equal”. In actual practice
today, 2205 duplex stainless steel usually contains 40 to 50 percent ferrite in solution annealed mill products. With
the relatively tight chemical composition of S32205, it would be remarkable if this steel were outside of this range.
Similarly, other duplex stainless steels with tightly specified chromium, molybdenum, nitrogen, and nickel ranges
(for example, 1.0, 0.5, 0.06, and 2.0% ranges, respectively) and well defined heat treatment ranges and cooling
requirements, may be expected to show very little variation in phase balance in the wrought mill products.
Metallographic methods provide the most direct and accurate evaluation of phase balance. Magnetic
methods, as discussed elsewhere, require correlation with metallographically evaluated standards of similar
geometry, composition, and thermal history.
Mill products. Determination of the actual volume fraction of austenite and ferrite by manual metallographic
examination using ASTM E 562 is a burdensome practice. It is burdensome because of the very fine microstructure
in a duplex stainless steel. It is best to determine the volume fraction by an automated procedure covered by ASTM
E 1245 with appropriate calibration standards. Measurement of the phase balance of mill products may be of little
value. The actual volume fractions will show little variation because the typical chemistries and heat treatments of
commercial heats fall within a very narrow range.
Phase balance measurement is not conservative with regard to its assessment of potential problems.
Metallographic examination, as discussed in Section 11, may reveal the presence of detrimental phases, but
determination of the austenite-ferrite phase balance will not detect intermetallic phases at the level known to reduce
impact toughness and corrosion resistance.
Weld qualification. Determination of the phase balance may be useful as part of weld qualification
procedures. It is especially useful for evaluating procedures that may involve rapid quenching of a weld created by
low heat input. Welding of very thin sections to very heavy sections, e.g., a tube-to-tubesheet weld, or attachment
of a sheet liner to a plate, or autogenous smoothing passes on a weld, may quench the weld and high temperature
HAZ so rapidly that the structure remains excessively ferritic. Resistance welds, made without provision for
retarded cooling may also exhibit a highly localized region of ferrite that affects the toughness and corrosion
resistance.
The only situation reported in which excessively austenitic welds occur is when the filler metal, usually one
associated with a flux-protected weld, is highly overalloyed with nickel to obtain better toughness. When such a
filler is used, it is assumed that there will be sufficient dilution from the base metal to obtain an increase in ferrite
above that of the undiluted filler metal. But when such a filler is used in a large weld where later passes have
minimal dilution from the base metal or from previous passes, the filler metal by itself can produce a weld metal
with low ferrite. Below 25% ferrite, there may be some concern that the material may not maintain the high level of
chloride stress corrosion cracking resistance associated with the duplex stainless steels. Therefore, it is appropriate
to select a filler metal with a sufficiently high minimum ferrite level when it is to be used in weld geometries likely
to produce regions of minimal dilution.
Metallographic examination
As demonstrated in ASTM A 923 Method A, and the research done in support of this procedure,
metallographic examination for intermetallic phases can be effective for assuring the quality of duplex stainless
steels. The examination is typically at 400-500X. It requires a high standard of metallographic preparation to
accurately reveal the austenite and ferrite. For revealing intermetallic phases, there is a choice of etchants, but 50%
sodium hydroxide electrolytic etch is one that has given consistent results. This etch stains the intermetallic phases
first, and then stains the ferrite. If the ferrite is stained without evidence of intermetallics, it may be concluded that
the intermetallics are present in a volume fraction less than that required to adversely affect the toughness or
corrosion resistance. This etch does not reveal the presence or absence of non-metallic phases, such as chromium
nitride.
Aside from the skill and experience required for metallography of duplex stainless steels, the greatest
problem associated with the use of metallographic examination is agreement on interpretation. Requirements that
the material be “free of” are not practical because they are not quantifiable. ASTM A 923 Method A addressed this
problem by making metallographic examination a screening test that may be used for acceptance but not rejection.
A “clean” structure is acceptable, while the “possibly affected” structure requires additional testing of toughness or
corrosion, Method B or Method C, respectively. Metallographic evaluation has the advantage of detecting any
centerline intermetallic phases. This is cause for rejection but recoverable with further heat treatment. Another
advantage of metallographic evaluation is that the acceptance criteria, i.e., the reference micrographs for similar
product forms of duplex stainless steels of different alloy contents are similar, even if the thermal history required to
produce the acceptable or unacceptable structures may differ.
Metallographic examination has the advantage of being relatively low cost, and for mill products, may be
performed in the basic package of tests that would be applied for each heat lot.
Magnetic measurements
Magnetic testing is an attractive concept because it is non-destructive and relatively inexpensive, and because
it can be applied to fabrications. However, the magnetic response of the ferrite phase in duplex stainless steels is a
function of many variables including the size, shape, orientation, and the chemical composition of the ferrite.
Optical metallographic correlation is difficult because the ferrite in a weld may be very fine, may be irregularly
shaped, and is difficult to etch in a way that reveals all the ferrite without overetching to the extent that
misidentification occurs. Round robin testing to determine volume percent ferrite in welds by magnetic
measurements resulted in unacceptable interlaboratory reproducibility. These difficulties lead the Welding Research
Council Subcommittee on Welding Stainless Steels to develop the concept of an arbitrary scale called the Ferrite
Number (FN) and subsequently extend it as high as 100 FN for use with duplex stainless steel welds. While volume
percent ferrite and FN approximately correlate at low FN’s, they do not at high FN’s.
The magnetic methods were developed for welds and even there, they have limitations. Kotecki has noted
that “the Ferrite Number system is quite adequate for weld metal, but is not usable in its present form for
measurements in the heat-affected zone of duplex ferritic-austenitic stainless steel welds, because the heat-affected
zone is too thin.”
In spite of being developed for weld metal measurements, magnetic methods have been applied in attempts to
determine the austenite-ferrite phase balance. The volume sampled by a magnetic probe is hemispherical in shape
and its size is a function of the strength of the magnet. As noted above, many variables influence magnetic
response. A magnetic probe which samples only a very small volume of material may give different responses at
the surface and at mid-thickness in a material, especially when that material has been straightened or flattened. As a
result magnetic methods cannot give an precise quantitative assessment of phase balance even when the instrument
is calibrated with known duplex stainless steel standards with a similar structure and range of ferrite.
Magnetic methods can not be used to detect precipitation of intermetallic phases such as sigma phase because
sigma phase is nonmagnetic, and the amount of sigma phase required for detrimental effect on toughness and
corrosion resistance is small.
Corrosion testing
Corrosion testing has the advantage of being intuitively related to performance in service. But it is important
to recognize that A 923 Practice C and the “modified G 48” test were never intended as a “fitness-for-service test”
for a specific application. These tests detect the formation of intermetallic phases to the extent that an important
engineering property, corrosion resistance, may have been compromised. The relationship of ferric chloride testing
to any service environment must be established independently.
The formation of intermetallic phases on the order of 1 vol. pct. will cause a significant lost of corrosion
resistance. It is useful to think of precipitation of intermetallic phases in terms of a sensitization, not to
intergranular corrosion but rather to pitting attack not limited to the grain boundaries. The low diffusion rates of the
elements in intermetallic phases are responsible for the pervasive nature of the precipitation. The effect may be as
large as a loss of 20°C or more in the critical pitting temperature. Even when the toughness of the steel is not of
particular concern in an application at somewhat elevated temperatures, the loss of corrosion resistance may be a
concern because that is usually the fundamental basis for selection.
Corrosion testing has the advantage of being almost insensitive to specimen location or orientation. If an
intermetallic phase is exposed anywhere on a ground surface, a ferric chloride test will find it.
ASTM A 923, Method C. It has been customary to specify corrosion testing in terms of a “modified ASTM
G 48 test”, but seldom has the modification been adequately detailed. G 48 is a laboratory tool, not an acceptance
test. It does not define a test temperature, test duration, or sample size and orientation. It does not produce a
numerical result, and its method of determining the presence or absence of pitting, needle scratching the surface
under magnification, is open to interpretation and controversy. ASTM A 923 Method B uses the same equipment as
G 48, but does not reference G 48. It defines for 2205 (S31803) that the test shall be at 25°C for 24 hours on a
sample with uniform surface at least equal to a 120-grit finish, or finer. Corrosion is measured by a weight loss that
is converted to a corrosion rate, so a standard specimen size is not necessity. The weight loss required is large
enough to be convenient to measure, but small enough to be associated with development of a single pit in 24 hours.
While G 48 allows ignoring corrosion on the edges of the sample, A 923 allows no such exception because it
evaluates for the presence of detrimental intermetallic phases, and these phases rarely emerge from the major faces
of the steel. Any acceptance of edge attack is a dangerous and non-conservative in duplex stainless steels testing.
The corrosion test has the advantage of being relatively rapid, requiring 24 hours plus sample preparation
time. It is not sensitive to size or orientation, provided that edge attack is not ignored. It is reproducible, with little
chance of a false indication for material without some detrimental phase and virtually certain of attack when a
detrimental phase is present.
Welded mill products. For mill products with a weld, such as pipe or fittings, it is necessary to reduce the
test temperature even when the product is annealed, and especially when the product is annealed in-line so that time
at temperature is short. While in-line annealing may eliminate detrimental phases, it is not long enough to assure
homogenization of the segregation that can occur in the solidification of an autogenous weld. Accordingly, A 923
set a test temperature of 22°C for S31803for welded mill products. At the time A 923 was written, all duplex mill
products were annealed. However, with the 1994 introduction of ASTM A 928 for duplex stainless steel pipe
welded with filler, it is possible to specify a condition, designated “HT-O”, where the product has not been
annealed. Recent reports indicate that unannealed, welded 2205 has difficulty passing a 22°C test. It is likely that A
923 will be changed to include unannealed products as a matter for agreement between seller and purchaser on a
case by case basis.
It is important to recognize that A 923 or “modified G 48” was never intended as a “fitness-for-service test”
for a specific application. It detects the formation of intermetallic phases to the extent that an important engineering
property, corrosion resistance, may have been compromised. The relationship of ferric chloride testing to any
service environment must be established independently.
Summary of applicable standard test specifications
The standard ASTM testing specifications typically applied to duplex stainless steels are listed and
characterized in Table II.
Table 2. Standard testing specifications applicable to duplex stainless steels
ASTM Testing Characteristics Comment

Specification Tested
A 923 Method A Metallographic examination Applicable to mill products. Useful for

for intermetallic phases acceptance, not rejection.
A 923 Method B Detection of intermetallic Applicable to mill products. Single specimen to
phases by impact testing. detect intermetallic phases. Acceptance criteria
must be revised for use with welds. Sensitive to
specimen location and orientation.
A 923 Method C Detection of intermetallic Generally applicable. Detects a particular
phases by corrosion testing. condition, not suitability for use.
G 48 Method A Detection of intermetallic Predecessor of A 923, Method C. G 48 is not
“Modified G 48” phases by corrosion testing. effective for acceptance unless sample
preparation, test parameters, evaluation
procedures, and acceptance criteria are defined
as in A 923.
A 370 Hardness Of relatively little use as higher hardness can be
associated with both favorable and unfavorable
metallurgical conditions.
A 370 Charpy tests (3 specimens ASME UHA 51 uses Charpy tests to
per test) demonstrate suitability for use when tested at
the minimum metal design temperature
(MMDT). Sensitive to location and orientation.
May not detect intermetallics.
E 562 Phase balance (manual) Useful for determination of volume fraction of
E 1245 Phase balance (automatic) austenite and ferrite when applied for multiple
fields at appropriate magnification. Sensitive to
orientation and metallographic technique.
Intermetallics not readily detected.
A 799 Determination of ferrite Not generally accurate without reference
content by instrument standards of similar geometry and history.
Does not detect sigma phase.
Appendix B. Commercially available duplex stainless steels
As is true for austenitic stainless steels, the duplex stainless steels are a family of grades with a range in
corrosion performance depending on their alloy content. The development of duplex stainless steels has continued
and modern duplex stainless steels can be divided into four groups:
• Lean duplex, such as 2304, which contain little or no deliberate Mo addition;

• 2205, the work horse grade accounting for more than 80% of duplex use;
• 25 Cr duplex, such as S32550 and S31260; and
• Superduplex, with 25-26 Cr and increased Mo and N, such as 2507.
The ASTM has now adopted 2304, 2205, 255, and 2507 as associated with S32304, S32205, S32550, and
S32750, respectively. These names, defining nominal chromium and nickel contents, are listed as “common name,
not a trademark, widely used, not associated with any one producer.” It is permitted to use these common names
interchangeably with the listed UNS designations when ordering material to the ASTM specification. In ASTM
specifications, 2205 is not associated with S31803.
1
Table III lists duplex stainless steels covered in ASTM specifications for plate, sheet, and bar products.
Although few of the duplex grades are patented, many of these grades are closely associated with particular steel
producers as proprietary developments. This situation is reflected in the naming of the grades. Only one duplex
stainless steel has an AISI designation, Type 329, a first generation duplex stainless steel without intentional
nitrogen addition. Many of the grades are commonly known by a number that reflects their typical chromium and
nickel contents, e.g., 2205 with 22% Cr and 5% Ni. These composition-based names are used by many producers
and are generally recognized. Some of these common names have been listed in the ASTM specifications for the
duplex grade, but trademarks of individual producers cannot be listed in ASTM specifications. However, all grades
in ASTM specifications are listed by their designations in the Unified Numbering System (UNS), as shown in Table
III. It is good practice to reference the UNS number when ordering to the standard specifications in order to get the
desired grade.
Table IV lists the ASTM specifications that are applicable to the duplex stainless steels, both wrought and
cast products. In general it is true that the duplex grades can be ordered as flat-rolled products, bar products,
forgings, and fittings which use the same ASTM specifications as used for Type 316L austenitic stainless steel, but
there are separate specifications to be used for duplex tubular products and for castings.
Many, but not all, of the duplex stainless steels are included in the ASME Boiler and Pressure Vessel Code,
either in the body of the Code or as Code cases. Users should be aware that the ASME has established additional
requirements for the qualification of duplex stainless steels and constructions of duplex stainless steels. As will be
discussed later, these requirements include Charpy tests (with specification of required lateral expansion) of the mill
products with sections greater than 3/8 inch (9.5 mm) at the Minimum Design Metal Temperature, and similar
qualification of procedures and fabrications. The qualification of a fabrication may require sample material from
the same mill product, something to be considered when ordering material. These requirements are detailed and
complex, and the fabricator and user are advised to review these requirements in detail before proceeding with an
application of duplex stainless steels that will require certification of compliance with the ASME Code.
The mechanical properties of the duplex stainless steels, as required by ASTM A 240, are compared to those
of Type 316L in Table V. The duplex stainless steels typically have yield strengths about twice those of the
common austenitic grades, but are less ductile. The duplex stainless steels also have substantially lower thermal
expansion coefficients and substantially higher thermal conductivities than Type 316L, with the values for duplex
stainless steels more closely resembling those for carbon steel.
Table 3. Composition of duplex stainless steels and representative austenitic stainless steelsa
UNS Typeb C Mn P S Si Cr Ni Mo N Cu Other

Number
Duplex Grades
S31200 ... 0.030 2.00 0.045 0.030 1.00 24.0-26.0 5.5-6.5 1.20-2.00 0.14-0.20 ... ...
S31260 ... 0.03 1.00 0.030 0.030 0.75 24.0-26.0 5.5-7.5 2.5-3.5 0.10-0.20 0.20-0.80 W 0.10-
0.20
S31803 … 0.030 2.00 0.030 0.020 1.00 21.0-23.0 4.5-6.5 2.5-3.5 0.08-0.20 ... ...
S32205 2205 0.030 2.00 0.030 0.020 1.00 22.0-23.0 4.5-6.5 3.0-3.5 0.14-0.20 ... ...
S32304 2304 0.030 2.00 0.040 0.030 1.00 21.5-24.5 3.0-5.5 0.05-0.60 0.05-0.20 0.05-0.60 ...
S32550 255 0.04 1.50 0.040 0.030 1.00 24.0-27.0 4.5-6.5 2.9-3.9 0.10-0.25 1.50-2.50 ...
S32750 2507 0.030 1.20 0.035 0.020 0.80 24.0-26.0 6.0-8.0 3.0-5.0 0.24-0.32 0.50 ...
S32760C ... 0.030 1.00 0.030 0.010 1.00 24.0-26.0 6.0-8.0 3.0-4.0 0.20-0.30 0.50-1.00 W 0.50-
1.00
S32900 329d 0.06 1.00 0.040 0.030 0.75 23.0-28.0 2.5-5.0 1.00-2.00 ... ... ...
S32950 ... 0.030 2.00 0.035 0.010 0.60 26.0-29.0 3.5-5.2 1.00-2.50 0.15-0.35 ... ...
Austenitic Grades
S30403 304Ld 0.030 2.00 0.045 0.030 0.75 18.0-20.0 8.0-10.5 … 0.10 … …
S31603 316Ld 0.030 2.00 0.045 0.030 0.75 16.0-18.0 10.0-14.0 2.00-3.00 0.10 … …
S31703 317Ld 0.030 2.00 0.045 0.030 0.75 18.0-20.0 11.0-15.0 3.0-4.0 0.10 … …
N08904 904L 0.020 2.00 0.045 0.035 1.00 19.0-23.0 23.0-28.0 4.0-5.0 0.10 1.0-2.0 …
S31254 … 0.020 1.00 0.030 0.010 0.80 19.5-20.5 17.5-18.5 6.0-6.5 0.18-0.22 0.50-1.00 …
N08367 … 0.030 2.00 0.040 0.030 1.00 20.0-22.0 23.5-25.5 6.0-7.0 0.18-0.25 0.75 …
N08926 … 0.020 2.00 0.030 0.010 0.50 19.0-21.0 24.0-26.0 6.0-7.0 0.15-0.25 0.50-1.50 …
S32654 … 0.020 2.0- 0.030 0.005 0.50 24.0-25.0 21.0-23.0 7.0-8.0 0.45-0.55 0.30-0.60 …
4.0
a Weight percent, maximum unless otherwise noted.
b Unless otherwise noted, a widely used common name, not a trademark, not associated with any one producer.
c Cr+3.3Mo+16N=40 min.
d AISI designation.
1,4,5
Table 4. ASTM Specifications Covering Duplex Stainless Steels
Product Form ASTM Specification

Plate, Sheet A 240
Bar Products A 276, A 479
Pipe A 790, A 928
Tubing A 789
Fittings A 815
Forgings A 182
Castings A 890, A 995
Testing A 923
1
Not all duplex grades are in each specification.
1
Table 5. ASTM A 240 Mechanical Test Requirements for Duplex Stainless Steels
Tensile Yield Elongation in 2 in.
UNS Strength, min Strength, min or 50 mm, min, % Hardness, max Cold Bend
Desig- Typ MPa ksi MPa ksi Brinell Rockwell C
nation e
S31603 316 485 70 170 25 40.0 217 95 (HRB) not required

L
S31200 ... 690 100 450 65 25.0 293 31 not required

S31260 ... 690 100 485 70 20.0 290 ... ...
S31803 … 620 90 450 65 25.0 293 31 not required

S32205 2205 620 90 450 65 25.0 293 31 not required
S32304 2304 600 87 400 58 25.0 290 32 not required
S32550 255 760 110 550 80 15.0 302 32 not required

S32750 2507 795 116 550 80 15.0 310 32 not required
S32760 ... 750 108 550 80 25.0 270 ... not required
S32900 329 620 90 485 70 15.0 269 28 not required

S32950 ... 690 100 485 70 15.0 293 32 not required
Appendix C. Frequently asked questions
The following questions reflect the experience of fabricators in the questions that most typically arise during
qualification of welding procedures. An attempt has been made to suggest answers, but in these practical matters,
there is a wide range of possibly “correct” answers. The answer given may not be applicable to all possible
situations.
1. Although it is recommended to use plasma torches for back gouging of the root and defect removal, can a
conventional carbon arc be used? What is a minimum grinding that should follow arc gouging in order to
remove heat-affected layer?
Carbon arc back gouging has been successfully used in the construction of 2205 duplex stainless steel
vessels, but care must be taken to minimize the heating and the potential for carbon contamination. When care is
taken in the back gouging procedure, the minimum grinding is not burdensome. It would be appropriate to perform
a weld procedure qualification in which the typically applied back gouging has been included as it will be used in
the practical construction.
2. What is maximum allowed thickness reduction resulting from cold forming before solution anneal/ water
quenching treatment would be required?
A precise answer to this question has not been developed. However, it has been common in construction of
2205 duplex stainless steel vessels to apply the same limits that are applied to carbon steels by the ASME Code.
This limitation, while possibly overly conservative, has not led to any problems in service. For vessels not being
constructed to ASME Code, significantly more aggressive deformation has been permitted, with no reports of
problems attributed to this cold worked condition.
3. What is a proper method of repairing small defects and metal tears on the process side (for example,
caused by knocking off of the pre-cut ladder supports and lifting lugs – usual method of removal)?
The repair procedure most typically satisfactory is to open the defect by grinding, if necessary, and then to
repair by GTAW with the typical “matching” filler, i.e., 2209 filler metal for 2205 fabrications. Because of the size
of the weld involved, it is unlikely that small defects or tears will lead to excessive time at temperature for these
repair welds. However, care must be taken to avoid too rapid cooling of the weld (with resulting excessive ferrite).
Slight warming of the metal under shielded conditions using the weld torch before the filler is introduced will
typically prevent too rapid cooling. Autogenous repairs are not normally recommended because control of
procedures is likely to be inconsistent and quality checks are inconvenient and uneconomical.
4. Excessive heat input may result from weld repair of the defect discovered by post-weld NDT. Should such
procedure be pre-qualified, and how?
It is appropriate to demonstrate that the weld repair has not damaged the material, i.e., to qualify the repair in
much the same way as the procedure was qualified. So fabricators have qualified reasonably anticipated repair
procedures in advance. Alternatively, the repair practice can be documented and simulated on a production runout
tab, with the usual production test plate procedures then being applied to the repaired weld.
5. What is the upper limit for weld heat input, provided base metal fully passed A-923 criteria?
Because the goal is to limit total time at temperature, it is generally better to complete a weld in fewer passes
with relatively higher heat input with adequate interpass cooling, than with many passes of lower heat input.
However, for relatively thin sections, the effectiveness of the workpiece itself in carrying away heat from the weld
may limit the heat input that is possible in each pass without excessive time at temperature (leading to precipitation
of intermetallic phases). The solidification of the duplex filler metals is not as prone to hot cracking as is a fully
austenitic solidification. It is not impossible to hot crack a duplex stainless steel during welding, but it is rare.
Duplex stainless steels can tolerate relatively high heat inputs without hot cracking because duplex stainless steels
have relatively low thermal expansion and high thermal conductivity.
6. Can heat input be allowed below the mentioned bottom value of 0.5 kJ/mm as long as the ferrite content
does not exceed 70% (for example, due to the over-alloying of the base and electrodes)?
Exceedingly low heat input is permitted, provided that the result is demonstrated to meet the usual
requirements for phase balance and corrosion resistance.
7. Does soda lime glass bead blasting provide an adequate finish for corrosive service, as an alternative to
pickling and what is the recommended surface profile range?
Whether or not a glass blasting will be sufficient for corrosive service will depend on the degree and nature
of the oxidized surface and the corrosivity of the service, including the tendency of the medium to adhere to the
surface of the steel. While a pickled surface provides corrosion resistance to the maximum capability of the grade, a
thoroughly blasted surface may be sufficient and economical. Scale and heat tint for the duplex stainless steels are
especially adherent and resistant to both mechanical and chemical removal.
8. What is the best way to prepare weld/HAZ specimens for A 923 Method C testing?
The specimen should be removed by the method least disruptive of the metal condition. Cold cutting is
recommended if possible. If a hot cutting method is applied, then there should be further cold cutting or grinding to
remove all material that was affected by the hot cutting. The surfaces tested should all be as ground without
pickling or other chemical treatment that might clear the surface of detrimental phases. A slight chamfering of the
edges is helpful, but the should not be substantial rounding off of the edges. The presence of burrs on the edges will
cause weight losses not related to the presence of intermetallic phases. Corrosion attack on the edges must be
included in the limiting acceptance criterion. “Modified G 48” procedures that permit disregarding of edge
corrosion are not correctly testing for the presence of detrimental intermetallic phases.
9. Is “modified G 48” testing the same thing as A 923 Method C?

ASTM G 48 Practice A and A 923 Method C are similar to the extent that they use similar equipment and
laboratory procedures. However, they are substantially different in their application. ASTM G 48 is a description
of laboratory procedure, but it does not specify the temperature of testing, the time of exposure, the technique of
assessing corrosion, and an acceptance criterion. The “modified G 48” test indicated that the individual ordering
specification was attempting to address these deficiencies, but few specification addressed all of them. ASTM A
923 Method C specifically addresses each of these issues, and provides a basis for acceptance of the duplex stainless
steels with regard to the absence of detrimental intermetallic phases.
One important difference is that G 48 permits the tester to disregard corrosion on the edges of the specimen.
This permission is totally inappropriate for use of the test to demonstrate the absence of intermetallic phases in
duplex stainless steels. It is unlikely that the intermetallic phases will occur in the faces of the plate or the faces of
the weld, but rather will occur in the interior of the metal. Therefore, incidents of pitting on the edges of the sample
should be considered indicative of a problem, and not ignored.
G 48 is usually a procedure performed at a series of temperatures, with the goal of identifying the critical
pitting temperature. Accordingly, the time of exposure and the inspection for pitting on the surface are designed to
detect subtle pitting initiation. The single test temperature for each grade in A 923 is chosen to be below the critical
pitting temperature for material without intermetallic phases, and above the critical pitting temperature for material
with intermetallic phases. The pitting, when it does occur, is readily visible. However, the weight loss is what is
measured in order to remove the potential for debate over visual interpretation. That weight loss is converted to a
corrosion rate in order to permit different sizes and geometries of specimens to respond to a single acceptance
criterion.
An important issue is the surface preparation of the sample. The goal of the test is to detect intermetallic
phases if present. Chemical treatment of the specimen surface (passivation or pickling) may reduce the exposure of
intermetallic phases in the surface and thereby cause the test not to detect the presence of intermetallic phases. The
specimen edges should be fine ground but not chemically treated for most effective use of the A 923 test. If there is
concern that the faces of the specimen may contribute to the weight loss, the appropriate specimen preparation is to
pickle the specimen before final grinding of the edges.
10. When you encounter a need to weld repair a structure of duplex stainless steel and you do not have a
detailed history of the welding during construction, how do you decide how much welding is safe? What
filler metal do you use?
The correct answer will depend on the nature of the weld, the conditions of application, and on the
application itself, particularly whether or not the structure was built to ASME Code, or is being used in a situation
of significant safety risk. The safest approach is to sample the fabrication weld and perform a qualification of the
proposed repair. However, this approach imposes extra costs and opens the necessity to repair also the position of
sampling. The value of good records in welding fabrication is amply demonstrated by this situation. It is
appropriate to consult metallurgical engineers before making the weld repair.
The problem, if there will be one, will most likely occur in the HAZ of the original fabrication welds. The
selection of the filler metal is unlikely to have any favorable effect on dealing with this part of the problem. The is
no reason that the filler metal should not be the same filler metal that would be used with the duplex stainless steel
in the original fabrication welds.
11. Are there any special problems in cleaning the heat tint of a duplex stainless steel?
Because of the relatively high chromium content and the relatively low thermal expansion of a duplex
stainless steel, the oxide scale is typically thin and highly resistant to removal. It is desirable to remove any heat tint
in order to get maximum corrosion resistance, but there are some applications where the process itself will remove
the heat tint. Grinding to clean bright metal is effective. Blasting can also be effective but, depending on the scale
and the blasting medium, may not be as effective as grinding for removing the oxide. Pickling, by solution or by
paste, is effective, but longer times or more aggressive pickling chemistries are required for duplex grades than are
typically required for austenitic grades.
Passivation, in the sense of removing free iron (from tooling contact, etc.), is no different than for austenitic
stainless steels. It is appropriate to confirm the effectiveness of a passivation treatment by testing such as that listed
in ASTM A 967.
12. When is post weld heat treatment beneficial, and what treatments should be used?
There are no heat treatments in the 315-1000° C (600-1800° F) range that are beneficial to duplex stainless
steels. Postweld stress relief heat treatments are used with steels that are capable of forming martensite, but duplex
stainless steels do not form martensite. The metallurgical condition of a duplex stainless steel will be severely
damaged if it is exposed to the stress relief treatment applied to a carbon or alloy steel (a consideration in dissimilar
welds).
If the duplex stainless steel for whatever reason is exposed to conditions that lead to the formation of
intermetallic phases, then the appropriate remedy is to heat treat the whole structure. The only heat treatment that
works for duplex stainless steel is a full anneal (above the minimum temperature listed in ASTM A 240, 1040° C
(1900° F) in the case of 2205) and quench. When the construction cannot be annealed and quenched, the only
remaining alternatives are to scrap the whole construction, or to cut out and replace the affected parts of the metal.
13. When is preheating useful or required?

Preheating the duplex stainless steel before welding is useful in two situations. If the part is damp, as from
condensation, heating uniformly to a maximum of about 95° C (200° F) will avoid the problems associated with
moisture in the weld. Preheating is one alternative for avoiding welds that are excessively ferritic as a result of too
rapid quenching. Examples include spot resistance welds, superficial surface repair, and welding of thin material to
heavy sections (sheet liners, tube-to-tubesheet welds). As with the suggested interpass temperature, 150° C (300°
F) is an appropriate maximum temperature for preheating.
14. What is the correct design for a runout tab?

The fact that the purpose of the runout tab is produce a sample of weld that is identical to the production weld
dictates the design of the tab. Ideally, the plate of the tab is from the same heat and thickness as the workpiece. It
should be of a size that will produce neither unusual heating or unusual cooling. It should be large enough to
readily supply the samples necessary for the qualification tests selected. Experience indicates that tabs from 6x6xt
to 12x12xt inches finished size have been satisfactory.
Sample material can be obtained from the plate itself when there are manways or nozzles to be cut, but this
source of samples may not always be available. When a bill-of-materials order is made for a large project
construction, with special sizes of plate being rolled, there may not always be off-cuts from the plates for the sample
material. It is a good idea to obtain the sample material with the purchase of the plate in order to assure the
availability of matching sample material.
15. How significant is the selection of the temperature for Charpy tests, comparing –40° C (-40° F) and the
ASME minimum design metal temperature?
ASTM A 923 and ASME UHA 51 have in common that they both use Charpy tests. However, the purpose
on the tests for the two procedures are quite different. The purpose of A 923 was to demonstrate that the heat
treatment applied to a duplex stainless steel mill product had eliminated the intermetallic phases. The Charpy test
was chosen because it was familiar to producer and user. As shown in the appendix of ASTM A 923, an acceptance
criterion of 40 ft-lb at –40° C (-40° F) was found to correlate with the appearance of the intermetallic phase in a
metallographic examination and a loss of corrosion resistance. Impact energy was selected as the acceptance
criterion because of its intuitive meaning and the fact that it is so readily measured in a familiar test procedure. A
923 was not intended to demonstrate suitability for use at this temperature. The test was chosen to demonstrate the
absence of the intermetallic phase. The high impact energy and low test temperature were necessary in order to get
a meaningful indicator for the extremely tough annealed mill product. A 923 states that its test procedures, and
especially the acceptance criteria, are not applicable to an as-welded structure.
In comparison, ASME UHA 51 is designed to demonstrate suitability for use. The temperature is minimum
design metal temperature, a factor of design specific to each installation. The standard test of three specimens is
performed using the lateral expansion measurement to confirm results. Impact strength well below 40 ft-lb is
accepted as suitable for use. It is applicable to the whole construction, whether base metal, weld metal, or HAZ.
Where appropriate, it is permitted to use the more demanding test conditions of ASTM A 923, but with the number
of specimens and measurements of both impact energy and lateral expansion, to qualify for ASME UHA 51, and so
reduce testing costs.
16. Why is 20 ft-lb impact energy sufficient for a weld when the specification for the plate requires 40 ft-lb at
–40° C (-40° F)?
The ASME has determined that 20 ft-lb is an adequate toughness for service in a particular class of
applications. This level of toughness is not high enough to correlate well with the observance of intermetallic phase
in the microstructure and the associated loss of corrosion resistance in a duplex stainless steel mill plate. The
duplex stainless steel plate structure is tough enough that it may still show significant impact energy even after
significant intermetallic phase formation. On the other hand, a weld metal may have toughness less than 40 ft-lb
even when no intermetallic phase is present.
17. Why is 25% ferrite enough for a weld, when higher ferrite content is specified for the base metal?
The base metal is specified with a composition that, for a fully annealed and quenched structure, will lead to
about 40 to 50% ferrite, essentially the equilibrium structure. This chemistry is found to return rapidly to almost
that balance after the thermal cycle that occurs in the HAZ during welding, retaining corrosion resistance and
toughness. It is known that the oxygen associated with flux shielding reduces the toughness of the weld metal.
Therefore, the compositions of the filler metal for flux-shielded welds have been adjusted to produce the highest
austenite that can be accepted while still retaining the benefits of the duplex structure. If there is significant
dilution from the base metal, then the weldment will have slightly more ferrite. The 25% ferrite represents the
minimum that will be achieved in there is essentially no dilution, as in a capping pass.
18. Is it necessary to water quench every time you heat treat a duplex stainless steel?
It is necessary to water quench after the final anneal of a mill product or of a constructed and heat-treated
component such as a head, fitting, or forging. However, it may be convenient to air cool the piece during
intermediate processing and then perform the final anneal and quench as a separate operation. The air cooled piece
will not have optimal toughness and corrosion resistance in that condition, but it is sufficient for further processing.
The part will be brought to maximum toughness and corrosion resistance by the final heat treatment with its water
quench.
19. Are there temperature limits, low and high, on the use of duplex stainless steels?
The toughness of the duplex stainless steel mill plate does not undergo an abrupt ductile-brittle transition.
Rather it decreases gradually from its high shelf energy to a very low impact energy as temperature decreases from
about ambient to temperatures in the range of –50 to –100° F. So the minimum application temperature is
determined in accordance with the tough of the duplex stainless steel. To date, there have been few applications
with minimum design metal temperature below –40° C (-40° F).
The maximum temperature for ASME Code applications is 315° C (600° F). The temperature was chose
because it represents the lowest temperature for the transformation curve for the 475° C (885° F) embrittlement
reaction. Below that temperature, the steel will not be embrittled by this reaction in many years of exposure. In
non-Code applications, it would be possible to consider use of 2205 in applications where there are limited
excursions in the range just slightly above the limiting temperature. However, the embrittling reaction is real and
exceptions to the 315° C (600° F) should not be undertaken without full knowledge and evaluation.
In the German TüV code, 2205 is limited to 280°C (535° F) for the unwelded condition except for seamless
tubing, and to 250° C (480° F) for the welded condition and for seamless tubes. In the TüV code, 2304 is limited to
300° C (570° F) in the unwelded condition, or when welded with matching filler, but is restricted to 250° C (480° F)
when welded with the “2209” filler typically applied as the matching filler metal for 2205.
20. How do the properties of duplex stainless steels affect wall thickness, thermal expansion, and heat
transfer in comparison to austenitic stainless steels?
Although it is generally correct to say that the yield strengths of the duplex stainless steels are twice that of
the common austenitic stainless steels, that relationship does not imply that the thickness of the duplex stainless
steel will be simply half that of the austenitic stainless steel in the same design. The higher strength of the duplex
grades is reflected in higher allowable design stresses in the ASME Code. Depending on the shape of the
construction, it is possible to reduce significantly the thickness of the material required when using duplex stainless
steel, an opportunity for cost savings.
The thermal expansion of a duplex stainless steel is between that of carbon steel and austenitic stainless
steels. This difference can be an advantage in structure with cyclic heating because there is less necessity to
accommodate the large expansions associated with the austenitic materials. On the other hand, using duplex
stainless steel within a construction of austenitic stainless steel may create problems when the duplex steel does not
expand to the same extent. The combination of high strength and lower expansion may mean that the duplex
stainless steel will impose high stresses at the point where it is joined to the austenitic structure.
Because the duplex stainless steel has a ferritic matrix, it is more efficient in heat transfer than the austenitic
stainless steels. This property, combined with the thinner material that results from economical use of the higher
strength of the duplex grades, can be used to significant advantage in heat transfer applications.

TAPPI TIP-0402-29 Duplex Welding Qualification

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

TAPPI TIP-0402-29 Duplex Welding Qualification

Uploaded by

Copyright:

Available Formats

TIP 0402-29

Qualification of welding procedures for duplex

General guidelines for welding duplex stainless steels

TIP Category: Automatically Periodically Reviewed (Five-year review)

Differences between duplex and austenitic stainless steels

Selection of starting material

Cleaning before welding

Heat input and interpass temperature

Postweld heat treatment

Phase balance in the weld

Dissimilar metal welds

Table 1. Welding consumables used for dissimilar metal welding.

Welding procedure qualification

Applicable welding methods

Guidelines for testing for procedure qualification

Starting material (mill products)

Welding procedure qualification testing

Testing for production weld test plates

Training of welders and inspectors

Digesters, Stainless Steel, Welding

Effective date of issue: June 22. 2001.

Ralph Davison, Chairman, Technical Marketing Resources

Appendix A. Testing requirements applicable to duplex stainless steels

• Conservative capability to address the property of interest;

Chemical composition requirements

Tensile testing requirements

Bend test requirements

Impact toughness requirements

Phase balance requirements

Summary of applicable standard test specifications

Table 2. Standard testing specifications applicable to duplex stainless steels

ASTM Testing Characteristics Comment

A 923 Method A Metallographic examination Applicable to mill products. Useful for

Appendix B. Commercially available duplex stainless steels

• Lean duplex, such as 2304, which contain little or no deliberate Mo addition;

UNS Typeb C Mn P S Si Cr Ni Mo N Cu Other

Product Form ASTM Specification

S31603 316 485 70 170 25 40.0 217 95 (HRB) not required

S31200 ... 690 100 450 65 25.0 293 31 not required

S31803 … 620 90 450 65 25.0 293 31 not required

S32304 2304 600 87 400 58 25.0 290 32 not required

S32550 255 760 110 550 80 15.0 302 32 not required

S32900 329 620 90 485 70 15.0 269 28 not required

Appendix C. Frequently asked questions

9. Is “modified G 48” testing the same thing as A 923 Method C?

13. When is preheating useful or required?

14. What is the correct design for a runout tab?

You might also like